Vector Databases
Vector databases are the backbone of the retrieval process in RAG systems. They store embeddings of document chunks and enable efficient similarity-based search.
Key Features of Vector Databases:
-
Index Types: How the database organizes data for efficient lookup.
-
Scale: How much data it can handle efficiently.
-
Hybrid Search: Ability to combine vector search with keyword or metadata filtering.
-
Cloud Ready: Whether it is designed to run easily in cloud environments.
-
Notes: Special features, ecosystem, or language support.
Popular Vector Databases:
| Name | Index Types | Scale | Hybrid Search | Cloud Ready | Notes |
|---|---|---|---|---|---|
| Faiss | IVF, HNSW | Local, fast | ❌ | ❌ | High-performance C++/Python library |
| Milvus | IVF, HNSW | Billion+ | ✅ | ✅ | Scalable and enterprise-grade |
| Qdrant | HNSW | Million+ | ✅ | ✅ | Rust/Python API, cloud-native |
| Weaviate | HNSW | ~10M+ | ✅ | ✅ | Strong semantic + metadata filters |
| Chroma | HNSW (simple) | Small-medium | ✅ | ✅ (basic) | Lightweight, easy local setup |
| SurrealDB | MTREE, HNSW | Billion+ | ✅ | ✅ | Multi-model DB, integrates SQL+Vec |
Summary of the Workflow
The full RAG pipeline includes:
-
Classifying the query to see if retrieval is needed.
-
Retrieving relevant documents using similarity search.
-
Reranking results for improved precision.
-
Repacking or organizing the information.
-
Summarizing it (optional, for length or clarity).
-
Generating the final answer using an LLM.
This modular workflow allows RAG systems to be flexible, extensible, and adaptable to various real-world needs.